International Journal of Image, Graphics and Signal Processing @ijigsp
Статьи журнала - International Journal of Image, Graphics and Signal Processing
Все статьи: 1157
Score Fusion of SIFT & SURF Descriptors for Face Recognition Using Wavelet Transforms
Статья научная
Automatic face recognition is a major research area in computer vision which aims to recognize human face without human intervention. Significant developments in this field have shown that in many face recognition applications the automated techniques outperform humans. The conventional Scale-Invariant Feature Transform (SIFT) and Speeded-Up Robust Features (SURF) are used in face recognition where they provide high performances. However, this performance can be improved further by transforming the input into different domains before applying SIFT and SURF algorithms. Hence, we apply Discrete Wavelet Transform (DWT) or Gabor Wavelet Transform (GWT) at the input face images, which provides denser and extra information to be used by the conventional SIFT or SURF algorithms. Matching scores of SIFT or SURF from each subimage is fused before making final decision. Simulations show that the proposed approaches based on wavelet transforms using SIFT or SURF provides very high performance compared to the conventional algorithms.
Бесплатно
Score-level-based face anti-spoofing system using handcrafted and deep learned characteristics
Статья научная
Recognition performance of biometric systems is affected through spoofing attacks made by fake identities. The focus of this paper is on presenting a new scheme based on score level and decision level fusion to monitor individuals in term of real and fake. The proposed fake detection scheme involve consideration of both handcrafted and deep learned techniques on face images to differentiate real and fake individuals. In this approach, convolutional neural network (CNN) and overlapped histograms of local binary patterns (OVLBP) methods is used to extract facial features of images. The produced matching scores provided by CNN and OVLBP then combined to form a fused score vector. Finally, the last decision on real and attack images is done by combining decisions of hybrid scheme using majority vote of CNN, OVLBP and their fused vector. Experimental results on public spoof databases such as Print-Attack and Replay-Attack face databases demonstrate the strength of the proposed anti-spoofing method for fake detection.
Бесплатно
Seamless Panoramic Image Stitching Based on Invariant Feature Detector and Image Blending
Статья научная
Image stitching is the method of creating a composite image from several images of the same scene. This paper addresses the issues of generating a seamless panoramic image from a series of photographs of the same scene by varying scale, orientation and illumination. A feature-based approach is proposed in this paper. Scale Invariant Feature Transform (SIFT) is used to detect key points in the image. SIFT is both a feature detector and descriptor. The common region between different images is identified by comparing the feature descriptors of each image. Brute-Force matcher with KNN algorithm is used for feature matching. The outliers in the matching features are eliminated by Random Sample Consensus (RANSAC) algorithm. To create seamless image, alpha blending operation is applied. Experiments are conducted on UDISD (Unsupervised Deep Image Stitching Data set). The overall performance of the proposed stitching method is evaluated based on metrics such as PSNR, SSIM, RMSE, MSE and UIQI, and the proposed stitching algorithm yields good result with seamless stitched image.
Бесплатно
Secure Data Transmission in Video Format Based on LSB and Huffman Coding
Статья научная
The growth of needing to transmit bit amount of data through the internet in secure format encourage the research for steganography technique, especially in video file. Stenographic technics in video format gives many advantages to transportation of important data because video files are a part of people’s daily life and the attackers can’t notice easily. The high embedding capacity of video file improves the popularity of video steganography among the various media types. Therefore, the simplest form but with many advantage of (Least significant bit) LSB, that is enforced with the high compression method of Huffman chunk coding method is proposed in this paper to embed data in video file in multi-step cryptography embedding schemes. The intension is to get more secure nature of the system and to get more embedding capacity system. The experiments are carried out with various sizes of video files and text file sizes are used to show the effectiveness of the proposed methods. The results manifest superior performance for proposed algorithm with the performance parameters like Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE) and Bit Error Rate (BER) are calculated to test the quality of stego video.
Бесплатно
Статья научная
In this paper a new type of information hiding skill in biomedical images is proposed through a combination of cryptography and digital watermarking to achieve the enhancement in confidential and authenticated data storage and secured transmission. Here patient's name and doctor's name are considered as patient's information which is encrypted using cryptography and embedded in the scan image of that patient through watermarking. RSA algorithm is used for encryption and higher order bit LSB replacement technique is used for embedding the information. The private keys are also embedded in the cover image to have better security and accurate recovery of the hidden information. The outcome of the proposed methodology shows that the hidden information doesn't affect the cover image and it can be recovered efficiently even from several noisy images. The strength of the proposed embedding scheme is also supported by several image quality matrices.
Бесплатно
Secured Lossy Color Image Compression Using Permutation and Predictions
Статья научная
Due to rapid growth in image sizes, an alternate of numerically lossless coding named visually lossless coding is considered to reduce storage size and lower data transmission. In this paper, a lossy compression method on encrypted color image is introduced with undetectable quality loss and high compression ratio. The proposed method includes the Xinpeng Zhang lossy compression [1], Hierarchical Oriented Prediction (HOP)[2], Uniform Quantization, Negative Sign Removal, Concatenation of 7-bit data and Huffman Compression. The encrypted image is divided into rigid and elastic parts. The Xinpeng Zhang elastic compression is applied on elastic part and HOP is applied on rigid part. This method is applied on different test cases and the results were evaluated. The experimental evidences suggest that, the proposed method has better coding performance than the existing encrypted image compressions, with 9.645 % reductions in bit rate and the eye perception is visually lossless.
Бесплатно
Segment-wise Quality Evaluation for Identification of Face Spoofing
Статья научная
Non-intrusive nature of the face-based recognition technology makes it more popular among hand held devices. Spoof detection in face-based recognition systems has been an important topic of the research in the last decade. Among several techniques available in the literature for liveness detection, image quality measure (IQM) based technique are particularly attractive due to their computational efficiency. In this paper, an approach based on segment-wise computation of image quality measures is proposed to improve the accuracy of detection. Two types of the non-overlapping segments are considered here: 1) rectangular segments of identical sizes, 2) segment based on neighborhood variance. It is found that both approaches exhibit better performance in comparison with other techniques without increasing too much computational complexity. The experiments are carried out with well-known Replay-Attack database to prove its robustness under different conditions.
Бесплатно
Segmentation and Counting of WBCs and RBCs from Microscopic Blood Sample Images
Статья научная
In the biomedicine field, blood cell analysis is the first step for diagnosis of many of the disease. The first test that is requested by a doctor is the CBC (Complete Blood cell Count). Microscopic image of blood stream contains three types of blood cells: Red Blood Cells (RBCs), White Blood Cells (WBCs) and platelets. Earlier counting of blood cell was done manually which was inaccurate and depends on operator's skill. Counting of blood cells using image processing provides cost effective and accurate result than manual counting. During the counting process, the splitting of clumped cell is the most challenging issue. This paper represents segmentation and counting of RBCs and WBCs from microscopic blood sample images. Segmentation is done using Otsu's thresholding and morphological operations. Counting of cells is done using geometric features of cells. RBCs contain clumped cells which make the task of counting of cells accurately very challenging. For counting of RBCs, two different methods are used: 1) Watershed segmentation 2) Circular Hough Transform. Comparison of both this method is shown for randomly selected images. The performance of counting methods is also analyzed by comparing it with results obtained by manual counts.
Бесплатно
Segmentation of Ancient Telugu Text Documents
Статья научная
OCR of ancient document images remains a challenging task till date. Scanning process itself introduces deformation of document images. Cleaning process of these document images will result in information loss. Segmentation contributes an invariance process in OCR. Complex scripts, like derivatives of Brahmi, encounter many problems in the segmentation process. Segmentation of meaningful units, (instead of isolated patterns), revealed interesting trends. A segmentation technique for the ancient Telugu document image into meaningful units is proposed. The topological features of the meaningful units within the script line are adopted as a basis, while segmenting the text line. Horizontal profile pattern is convolved with Gaussian kernel. The statistical properties of meaningful units are explored by extensively analyzing the geometrical patterns of the meaningful unit. The efficiency of the proposed algorithm involving segmentation process is found to be 73.5% for the case of uncleaned document images.
Бесплатно
Segmentation of Mammogram Images Using Optimized Kernel Fuzzy AGCWD Based Level Set Method
Статья научная
Image enhancement technology is widely used to improve images and help radiologists make more accurate cancer diagnoses. In this research work presents an integrating approach for contrast enhancement followed by the segmentation of breast cancer from the mammogram images. The proposed method has been effectively utilized the three different algorithms such as differential Evolution (DE) Algorithm, Kernel Based Fuzzy C Means (KFCM) Clustering and Cuckoo Search Optimization (CSO) algorithm. Here an integrating approach introduced, called Optimized Kernel Fuzzy Adaptive Gamma Correction with Weighed Distribution (OKF-AGCWD) based Level Set Method. The performance of proposed method is enhanced over existing level set methods such as image and vision computing (IVC)-2010, IVC-2013, and Expert Systems with Applications (ESA) 2021. The performance metric parameters like F1_score, Sensitivity, Specificity and accuracy are considered to assess the quality of different methods. The simulation was performed on 16 distinct images from the RIDER mammography database. The experimental results were compared with existing level set approaches such as image and vision computing (IVC)2010, IVC2013 and expert systems and applications (ESA)2021 with respect to OKF-AGCWD. The proposed OKF-AGCWD with contextual level set method (CLSM) minimizes boundary leakage problem of mammogram segmented image and improves segmentation accuracy.
Бесплатно
Segmentation of Medical X-ray Bone Image Using Different Image Processing Techniques
Статья научная
Accurate medical image processing plays a crucial role in several clinical diagnoses by assisting physicians in timely treatment of wounds and mishaps. Medical doctors in the hospitals generally rely on examining bone x-ray images based on their expertise, knowledge and past experiences in determining whether a fracture exist in bone or not. Nevertheless, majority of fractures identification methods using X-rays in the hospitals is beyond human understanding due to variation in different attributes of fracture and complication of bone organization thereby making it difficult for doctors to correctly diagnose and proffer adequate treatment to patient ailments. The need for robust diagnostic image processing techniques for image segmentation for different bone structures cannot be overemphasized. This research implemented different image segmentation techniques on a bone x-ray image in order to identify the most efficient for timely medical diagnosis. Also, the strength and weaknesses of the diverse segmentation techniques were also identified. This will empowered researchers with appropriate knowledge needed to improve and build better image segmentation models which doctors can use in handling complex medical image processing problems. Also, miss rate in bone X-rays that contains multiple abnormalities can be lowered by using appropriate image segmentation techniques thereby improving some of the labor intensive work of medical personnel during bone diagnosis. MATLAB 9.7.0 programing tool was used for the implementation of the work. The results of X-ray bone segmentation revealed that active contour model using snake model showed the best performance in detecting boundaries and contours of regions of interest when used in segmenting Femur bone image than the other medical image segmentation approaches implemented in the work.
Бесплатно
Segmentation of Pre-processed Medical Images: An Approach Based on Range Filter
Статья научная
Medical image segmentation is a frequent processing step. Medical images are suffering from unrelated article and strong speckle noise. In this paper, we propose an approach to remove special markings such as arrow symbols and printed text along with medical image segmentation using range filter. The special markings are extracted using Sobel edge detection technique and then the intensity values of the detected markings are substituted by the intensity values of their corresponding neighborhood pixels. Next, three different image enhancement techniques are utilized to remove strong speckle noise as well enhance the weak boundaries of medical images. Finally range filter is applied to segment the texture content of different modalities of medical image. Experiment is conducted on ImageCLEF2010 database. Results show the efficacy of our proposed approach which lead to have precise content based medical image classification and retrieval systems.
Бесплатно
Segmentation of abnormal blood cells for biomedical diagnostic aid
Статья научная
The aim of our work is to obtain a maximum rate of recognition of abnormal (cancerous) blood cells. We propose the development of a system based on k-means methods, after an RGB channel decomposition by applying the algorithm which can segment our microscopic medical images. It turns out that the proposed system shows better segmentation and classification for the identification and detection of leukemia. The experimental results obtained are very encouraging, which helps hematologists to monitor the evolution of cancerous blood cells and make a good diagnosis.
Бесплатно
Segmentation of the herniated intervertebral discs
Статья научная
This paper presents two segmentation algorithms for MR spine image segmentation helping in on time diagnosis of the spine hernia and surgical intervention whenever required. One is level set segmentation and another one is watershed segmentation algorithm. Both of these methods have been widely used before (Aslan, Farag, Arnold and Xiang, 2011) (Pan, et al., 2013) (Silvia, España, Antonio, Estanislao , and David, 2015) (Erdil, Argunşah, Ünay and Çetin, 2013) (Claudia. Et al, 2007). In our approach we have used the concept of variational level set method along with a signed distance function and is compared with the watershed segmentation which we have already implemented before on a different dataset (Hashia, Mir, 2014). In order to check the efficacy of the algorithm it is again implemented in this paper on the sagittal T2-weighted MR images of the spine. It can be seen that both these methods can become very much valuable to help the radiologists with the on time segmentation of the vertebral bodies as well as of the intervertebral disks with relatively much less effort. They both are later compared with the golden standard using dice and jaccard coefficients.
Бесплатно
Статья научная
Image segmentation plays the significant roles in image processing, computer vision and as well as in pattern recognition. The Segmentation process subdivides an image into its constituent parts or objects, such that level of subdivision depends on the problem to be solved. The aim of image segmentation is partitioning an image within homogeneous regions that are significantly meaningful concerning some characteristics like intensity or texture. Based on clustering, a large number of researches have been done in the area of image segmentation. This paper presents an efficient image segmentation method in which the self organizing feature map (SOFM) is used for initial segmentation. After the initial segmentation, the segmented image is used by the K-means algorithm for further segmentation. Finally, the procedures for automatic splitting and merging the cluster are applied to obtain the appropriate number of segments in segmented image and as well as better segmented results. For analyzing the performance, we calculate the statistical measure named as Davies-Bouldin index (DB-index). The observation shows that, this method gives the better segmented results compared with K-Means algorithm, linear discriminant analysis (LDA) and K-Means based image segmentation method and also SOFM and K-Means based image segmentation approach.
Бесплатно
Self-supervised Model Based on Masked Autoencoders Advance CT Scans Classification
Статья научная
The coronavirus pandemic has been going on since the year 2019, and the trend is still not abating. Therefore, it is particularly important to classify medical CT scans to assist in medical diagnosis. At present, Supervised Deep Learning algorithms have made a great success in the classification task of medical CT scans, but medical image datasets often require professional image annotation, and many research datasets are not publicly available. To solve this problem, this paper is inspired by the self-supervised learning algorithm MAE and uses the MAE model pre-trained on ImageNet to perform transfer learning on CT Scans dataset. This method improves the generalization performance of the model and avoids the risk of overfitting on small datasets. Through extensive experiments on the COVID-CT dataset and the SARS-CoV-2 dataset, we compare the SSL-based method in this paper with other state-of-the-art supervised learning-based pretraining methods. Experimental results show that our method improves the generalization performance of the model more effectively and avoids the risk of overfitting on small datasets. The model achieved almost the same accuracy as supervised learning on both test datasets. Finally, ablation experiments aim to fully demonstrate the effectiveness of our method and how it works.
Бесплатно
Semantic Segmentation of Tuberculosis Bacilli from Microscopic Sputum Smear Images Using TransUNet
Статья научная
According to the World Health Organization (WHO) touchstones of 2022, Tuberculosis is the second dominant disease after COVID-19. Around one-fourth of the comprehensive population is ascertained to have tuberculosis. Timely detection and prevention of tuberculosis is a must to overcome its harmful effects. The method most often used in ascertaining whether a patient has tuberculosis, is examining his or her sputum sample. In the process, the isolation of the bacilli is done manually, and hence it is prone to error. Segmentation illustrates and enlightens objects or particles within an image, thus extracting the Region of Interest (ROI). The contemplated study uses TransUNet architecture to segment tuberculosis bacilli from sputum images to increase diagnostic accuracy and performance. The attention mechanism used in the TransUNet model helps to identify the spatial hierarchies present in image. It is an extremely tough task for naive or traditional segmentation algorithms to deal with the inherent complexity of sputum images. Hence, this study introduces an approach to capture the intrinsic features and dependencies needed to segment mycobacterium or TB bacilli by leveraging the TransUNet model. The model achieved an average Dice Score of 92.795%, a mean Intersection over Union (IoU) of 88.845%, and a segmentation accuracy of 99.19% on the Mosaic and Ziehl-Neelsen datasets. These results surpassed several existing state-of-art methods like UNet, clustering and thresholding, depicting the superior capability of TransUNet in segmenting the TB bacilli. It deepens the potentiality of transformer-based CNN models, especially TransUNet, for improving the diagnosis of tuberculosis and supporting disease management.
Бесплатно
Semi-fragile Image Watermarking Algorithm Based on Region-Segmentation
Статья научная
In order to improve the invisibility and the robustness of semi-fragile watermarking, the paper first brings up the idea which embedding watermark based on the attacks’ characteristic, brings forward the region segmentation operator and the image-segmentation embedding method, puts forward the characteristic and its representation in DWT transform domain based on visual features model, and brings forward the quantized central limit theorem which applies to adjusting the coefficients in general transform domain. These all make semi-fragile watermarking embedded through dynamic quantization achieve the greatest robustness. It gives wavelet transform domain coefficient redressal operator and the best restoration probability of the pixel value adjusting in experiments when the images were under attack. It leads up to a better invisibility of carrier image, a better robustness to the image processing, such as JPEG compression, noise adding, filtering, and the larger amount of embedded information. What’s more, it can ascertain the position of vicious attack exactly.
Бесплатно
Статья научная
The study is devoted to the analysis of public sentiment towards Ukrainian political figures based on comments on social media, in particular, YouTube and Twitter. The work aims to identify differences in the perception of political leaders and to understand how the platform affects the tone of statements. The main research question is to determine how public opinion about politicians in Ukraine differs between YouTube and Twitter during the full-scale war. To do this, a corpus of comments and tweets from 2022 to 2023 was collected, which went through pre-processing stages (including cleaning up slang and spelling mistakes). The article presents the results of a comprehensive analysis of public opinion on five public figures of Ukraine (S. Prytula, P. Poroshenko, V. Zelensky, S. Sternenko, A. Yermak) based on data from the social networks YouTube and Twitter. For data collection, the YouTube Data API and the Apify platform were used, a corpus of Ukrainian-language comments and tweets was collected and processed, which went through the stages of purification, normalisation and lemmatisation, taking into account slang, surzhyk and spelling mistakes. The sentiment analysis model, built on the basis of multilingual-e5-base embeddings and the XGBClassifier algorithm, showed an accuracy of 89.4%, macro-F1 of 88.7%, and a weighted F1 of 89.1%. Sentiment distribution analysis revealed that, on average, 42% of messages were positive, 36% were negative, and 22% were neutral. Twitter had a higher share of negative statements (up to 40%), while YouTube had a predominance of positive sentiment (up to 47%). The results indicate differences in the perception of public figures on different platforms and confirm the effectiveness of the developed approach for the Ukrainian-speaking segment of social networks. The results indicate significant differences in sentiment distribution: comments on YouTube are more likely to be marked by emotional intensity and harshness. At the same time, Twitter exhibits a more concise but no less polarised discourse. One of the reasons for this difference may be the difference in the format of the platforms, their audience, and the speed of content distribution. Further research should take into account the impact of user demographic biases, as well as the activity of bots or coordinated campaigns that can change the perception of public opinion. The practical significance of the study lies in the fact that its results can be used by politicians, journalists, and public figures to better understand the mood of society, predict reactions to political events, and build more effective communication. At the same time, it is worth noting that there are limitations: automated sentiment analysis has difficulty detecting sarcasm, irony, or context-sensitive meanings, which can affect the Accuracy of the results. In addition, the study takes into account the ethical aspects of data collection and analysis: only publicly available comments were used, without interference in the private sphere of users. There are possible risks of abuse of such technologies, and the need for responsible application of the findings is emphasised.
Бесплатно
Serial digital color image watermarking using composite scheme
Статья научная
Digital watermarking is one of the ways to have Copyright protection for digital information. The digital watermarking scheme used for watermark embedding has to satisfy robustness property to ensure the security of the secret information hidden. The scheme presented here will support the above said statement significantly. We propose here the scheme as composition of both blind and non-blind digital watermarking technique in a process of serial watermarking. A secret binary image is embedded in the first cover image to get first watermarked image by using blind watermarking technique. Then this first watermarked image is again embedded into second cover image to get serial watermarked image using non-blind watermarking technique. To extract secret binary image, first non-blind watermark extraction technique and then blind watermark extraction techniques are used. From this composite approach and serial watermark embedding procedure, we achieved considerable fidelity and robustness against - Rotation, JPEG compression and for noises Salt & pepper, Gaussian, Speckle, Poisson and multiple noises.
Бесплатно
